Regnerative Braking

[mechanix]

I have a question about regenerative braking - People are always talking about how if you're pushing the limits of voltage for a controller and regen kicks in, you could do some damage (blow caps, fets, etc...). I did a search on this and nothing came up, so I thought I'd ask you guys.

My question is:

How does regen work? I know that the motor is being utilized as a generator, but how is the output voltage determined? Is it a function of how fast it is spinning? Does the controller regulate the voltage that the batteries will see? The reason I'm asking is because I'm planning a build that would most likely have 24s Lifepo4 (~86.4V fully charged, 79.2 nominal) and I'm wondering how close to 100V (the rating of my controller) the voltage will spike when regen kicks in on a full battery. I'm not trying to push the limits of the equipment, my goal is for this to be a very reliable daily commuter. What if I had 28s (100.8v, or 92.4V nominal)?

-Joel

 

[wesnewell]

I've been running regen on 24s lipo (100.8V charged) for 4.5 years without a problems on a cheap 72V controller with 100V caps and fets. Works great.

 

[Alan B]

Hi Joel,

In Regen mode the controller uses the magnetic field in the motor coils, together with pulsing the FETs in the controller in a certain way, to convert the voltage being generated by the motor upwards (called boost conversion) so that current may flow back into the battery, creating drag on the motor and charging the battery.

This technique increases the motor's generated voltage up to battery voltage plus enough extra to overcome the losses in the wiring and the controller electronics.

The case wesnewell mentions is fairly extreme and with 24S lipo is exceeding the ratings of the FETs and Caps in his controller, but at least in that one case they seem to be working. As an engineer I would not recommend exceeding or even approaching the ratings like that, but it is his choice.

For good reliable operation it is generally best to stay below 80% of the ratings.

The low cost controllers don't do a very good job of limiting the maximum voltage. More sophisticated controllers will have adjustable voltage and current limits on regen to avoid overcharging the battery or damaging the electronics, and they will also have variable regen and will even apply power to allow electric braking at low speeds, where regen alone becomes insufficient.

 

[justin_le]

How does regen work?

At an electronics level, the best answer to this is to look up "DC-DC Boost Converter". Regen is exactly the same thing as this, with the motor controller boosting the back-emf voltage of the motor up to the battery supply voltage.

I know that the motor is being utilized as a generator, but how is the output voltage determined? Is it a function of how fast it is spinning?

No, not at all. The output voltage during regen is determined entirely by your battery pack. The controller can regulate the current or power flowing into the battery during regen, but the voltage present on the battery while this power flows in is just a function of the battery terminals. Say you have a battery with an internal resistance of 0.2 ohms and it's currently sitting at 52.5V. If you do intensive regen and put 20 amps into the pack, it would be at (52.5V * 20*0.2) = 56.5V. If your controller was doing just a modest 5A of regen current, then the voltage during regen would be 53.5V etc.

Most controllers will have a maximum regen voltage, and if the battery voltage exceeds this then the controller will stop doing regenerative braking. That is a spec that you should be able to find on your particular motor controller. If a controller doesn't have a maximum regen cutoff, and you do regen and then either unplug the battery or have the battery BMS circuit trip, then the voltage will keep on increasing as the regen current has nowhere to go until something gives (usually the mosfets).

Does the controller regulate the voltage that the batteries will see? The reason I'm asking is because I'm planning a build that would most likely have 24s Lifepo4 (~86.4V fully charged, 79.2 nominal) and I'm wondering how close to 100V (the rating of my controller) the voltage will spike when regen kicks in on a full battery.

To answer that, you will need to ask the vendor who sold you the controller to find out the max regen voltage. Alternately, you can find this out fairly easily by hooking a voltmeter to the controller, run the motor up to full RPM with the battery attached, then quickly unplug the battery and apply the ebrakes. With no battery to absorb the regen energy, the voltage on the controller should increase to the max regen voltage setting and then stay there, while the wheel slowly spins down.

 

[Warren]

Justin,

I'd love to hear your take on my regen experience. On my cargo bike I am running this 18 fet 4110, 36-75V Infineon controller.

http://em3ev.com/store/index.php?route=product/product&path=41_38&product_id=83

At high speed, if I engage the regen button and hold it, the regen will rise as high as 980 negative watts (the highest negative number I have noticed). If I am going fast enough, or down a steep enough hill, at 130 pounds, the bike still has lots of potential energy to burn off, and I start to see positive watts reading in the 1000-1600 watt range. All while still decelerating hard. In talking to Bill Bushnell, he said it must be "plug braking"...actually using battery energy to slow the bike. Since that conversation, I tend to push the regen button intermittently to avoid using battery energy to slow down, unless I am in traffic, and need to slow rapidly. Does Bill's assessment sound right?

Thanks, Warren

 

[wesnewell]

That isn't regen braking, it's electronic braking that uses battery power to reverse the phases to stop the motor. It will bring my bike to a complete stop. Regen braking won't do that. My old controller has both. Grounding the low brake line engages regen braking that will slow the bike to about 5-10 mph and charge the battery before disengaging. Connecting positive battery to the hi-brake line causes electronic braking that uses battery power and will bring the bike to a complete stop. And it's a lot stronger braking power than regen. Note that LVC affects both braking types. If LVC is set too low, then neither braking type will function until the battery level falls withing a range for it it to function. My default LVC was ~62V on my 72V controller and neither braking type would work with a battery voltage above ~78V. After raising LVC to ~86V it would work at much higher voltages.

 

[LewTwo]

Grounding the low brake line engages regen braking that will slow the bike to about 5 mph and charge the battery before disengaging.
Connecting pos battery to the hi-brake line causes electronic braking that uses battery power and will bring the bike to a complete stop.

Thank thee. Best succinct logical explanation I have seen for hi vs lo brake wires.

 

[Warren]

I don't have any low brake line, or high brake line. Don't know what that is. The e-brake plug, from the brake levers, is plugged into the e-brake connector on my CA. It just cuts throttle when the brake levers are pulled. The regen button is plugged into the regen plug on the controller. It completes the regen circuit in the controller, giving full regen, with no modulation. But beyond a certain point it goes to plug braking/electronic braking. It will not bring the bike to a stop. It pulls it down to 12 mph, and drops out. At that point the mechanical brakes work fine, even on a 130 pound bike. Real, progressive regen, to a stop, like Justin now offers, would be sweet.

 

[justin_le]

Justin,

I'd love to hear your take on my regen experience. On my cargo bike I am running this 18 fet 4110, 36-75V Infineon controller.

http://em3ev.com/store/index.php?route=product/product&path=41_38&product_id=83

At high speed, if I engage the regen button and hold it, the regen will rise as high as 980 negative watts (the highest negative number I have noticed). If I am going fast enough, or down a steep enough hill, at 130 pounds, the bike still has lots of potential energy to burn off, and I start to see positive watts reading in the 1000-1600 watt range.

I think the issue here is that you have just lost the negative symbol (-) part of the display on the CA screen from the addition of your extra digit. Switch over to amps rather than watts on the main screen and you'll see it is negative all the while.

You will see plug braking with certain controllers at very low speeds but not usually with the infineon style controllers, on them the regen intensity reduces as you come to a stop so it never enters this domain. Going fast and doing regen will always involve putting energy back into the pack.

 

[Warren]

Justin,

Thanks so much for the reply. Sounds right. I will give this a try the next time I get out. Cold and snow flurries Monday, but maybe Tuesday or wednesday. You da man!